AI

Question 1

What is a great saying about AI & morality?

Answer 1

• AI relies on black-and-white rules, whereas morality is famously a morass of endless grey (and if humans categorised humans can breed great injustice, machines categorising humans will be no different)

Question 2

How many jobs will be lost to AI?

Answer 2

• According to a study by Oxford Economics, up to 20 million manufacturing jobs will get lost to robots by 2030 all across the world, as such algorithmic tasks, which involve predictable parameters and pre-set steps, can easily be defined as machine-readable code

Question 3

How would AI affect jobs? (Goldman Sachs)

Answer 3

• After the generative AI boom in 2023, Goldman Sachs updated the estimate to two-thirds of jobs in Europe and the US being exposed to at least some AI automation, and one-quarter of jobs being entirely performable by AI

Question 4

How does AI threaten white-collar professions by moving towards higher-order cognitive and creative tasks usually associated with humans, like perception, planning and perform-solving

Answer 4

• As early as 2017, CheXNet, a neural network developed by Stanford to interpret chest X-rays, was beating human doctors, experts, and radiologists for accuracy in diagnosing 14 different conditions. In 2020, it was further fine-tuned to detect Covid-19 by zeroing in on pneumonia symptoms, boasting a reported 99% accuracy

Question 5

Data and IRL example of consequences caused “when AI systems learn from data which may be unbalanced or reflect discrimination, resulting in outputs which have discriminatory effects on people based on their gender, race, age, health, religion, disability, sexual orientation or other characteristics”

Answer 5

• In an extensive government study by the US National Institute of Standards and Technology (NIST), federal researchers have found widespread evidence of racial bias in nearly 200 facial recognition algorithms. Racial minorities were found to be magnitudes likelier to be misidentified than whites, sometimes by a factor of 10 or even 100
• (The use of face recognition softwares is growing rapidly in law enforcements, border control, and other applications throughout society)
• Resulted in wrongful arrests like in the cases of Michael Oliver, Nijeer Parks, and Robert Williams — three African-American men whose lives were derailed after they were wrongly arrested and subjected to traumatic legal battles after being misidentified by facial recognition softwares

Question 6

How are AI bias prevalent in the healthcare system?

Answer 6

• A major health care risk algorithm used across several U.S. health systems was a found to exhibit bias by prioritising healthier white patients over sicker black patients for additional care management because it was trained on cost data, not care needs
• (Even benign applications in healthcare to identify bumps, lumps, and mumps risk mishap, as all conjectures involve error bars)

Question 7

Regulations on AI (EU)

Answer 7

• In 2024, the EU adopted a new AI Act that defines several risk levels for AI. Projects posing an “unacceptable risk,” like biometric identification systems, or those that categories people for social scoring or cognitively manipulate vulnerable groups like kids, are straight-up banned, while those presenting a “high risk” are subject to stiff laws

Question 8

Why is there a global AI race?

Answer 8

• (Revolutionary technologies in AI will bestow upon the wielded an insurmountable advantage — like how the UK rose to be a global superpower following the Second Industrial Revolution, and the US rode the Third Industrial Revolution to emerge as the wealthiest nation ever, whoever takes advantage of the Fourth will rule the world )

Question 9

How did Ukraine become a testing ground for AI?

Answer 9

• Ukraine has been described as “the world’s tech R&D lab” and “Mil-Tech Valley” as technology companies, many focused on autonomous drones, have set up shop in Kyiv since the war. As the future of warfare is being beta tested on the ground in Ukraine, the results will have global ramifications
• War has always driven innovation, from the crossbow to the atomic bomb. In conflicts wages with software and AI, where more military decisions are likely to be handed off to algorithms, the prospects for escalations are sky-high

Question 10

Examples of AI used for war purposes

Answer 10

• Senior officials in Ukraine revealed how they have been using dozens of domestically made AI-augmented systems for its drones to reach targets on the battlefield without being piloted, allowing it to remain effective in areas protected by extensive jamming